>>> A LIFE IN TECH: ENGELBART’S ACHIEVEMENTS
In the early 1970s some key members of Engelbart’s ARC team left to join Xerox PARC, and by 1975, when ARPA funding dried up, only Engelbart was left. SRI sold NLS to a phone networking company called Tymshare, where Engelbart continued working on it in obscurity. Steven Levy interviewed him there in the early 80s and found him disappointed. From the late 80s onwards, he received some measure of recognition and several awards, such as the IEEE Computer Pioneer Award in 1992, and the prestigious Turing Award in 1997, but still dreamed of seeing his system widely adopted — not just the tools but the intellect–augmenting organisational structures and paradigms.
His daughter Christina said he hated being pigeonholed as the inventor of the mouse. He was frustrated that of all his innovations only the simplest, the mouse, had been taken up. He was disparaging of the Mac’s focus on ease–of–use and WYSIWYG — “what you see is all you get,” he complained.
Engelbart died at the age of 88 in July 2013. Christina still runs the Doug Engelbart Institute, dedicated to his vision of augmenting human intellect. Find it at dougengelbart.org. away over a leased line link–up, but it wasn’t the computer he was demonstrating. It was interactive computing, a radical new way of using a computer, in contrast to the standard number–crunching.
The presentation is now known as “the mother of all demos.” It was given that name in a book by tech journalist Steven Levy published in 1994; the common Middle Eastern “mother of all...” idiom had become current in the West when Saddam Hussein threatened “the mother of all battles” in a speech during the first Gulf War in 1991. The book was titled Insanely Great: The Life and Times of Macintosh, the Computer That Changed Everything. In it, Levy made the connection between Engelbart’s demo and the way the Mac worked. It was, in numerous tangible ways, the prototype of the Macintosh. Vannev ar Bush and the memex Not that Engelbart’s ideas sprang out of thin air: his entire career had been inspired by the ideas of Vannevar Bush, who published an article in 1945 in The Atlantic magazine entitled As We May Think. The editor of The Atlantic explained in an introduction, “As Director of the Office of Scientific Research and Development, Dr Vannevar Bush has coordinated the activities of some six thousand leading American
scientists in the application of science to warfare.” Looking forward to when the fighting has ceased, “[Bush] urges that men of science should then turn to the massive task of making more accessible our bewildering store of knowledge.”
In his article, Bush pictured a time when “The Encyclopedia Britannica could be reduced to the volume of a matchbox. A library of a million volumes could be compressed into one end of a desk.” Essentially, he conceived of vast amounts of information being digitized, as we’d describe it today, and extrapolated from the very limited technologies of the day to speculate how this might be done — for example, he cleverly postulated a method of “dry photography” analogous to the way a television screen works. The desk containing all this information, he envisaged as “a sort of mechanized private file and library,” which he dubbed a “memex”. “While it can presumably be operated from a distance, it is primarily the piece of furniture at which [somebody] works. On the top are slanting translucent screens, on which material can be projected for convenient reading. There is a keyboard, and sets of buttons and levers.
“Most of the memex contents are purchased on microfilm ready for insertion. Books of all sorts, pictures, current periodicals, newspapers, are thus obtained and dropped into place. Business correspondence takes the same path. And there is provision for direct entry.” The user would then navigate through the stored material by tapping a code on the keyboard, and use levers (or a joystick much like the controller on a modern microfilm reader) to move forward or backwards, to find the desired information and then display it.
More than this: the user would be able to add annotations and links to other material, building “a trail of his interest through the maze of materials available.” With a trail saved, it could then be followed again later and even shared with other people to follow. In this way, the vast mass of information could be navigated through, selected, interpreted and understood much more readily. Augmenting human intellect Doug Engelbart read Bush’s essay as a young radar technician stationed in the Pacific. Some years later, it inspired his choice of career — or, as he called it himself, his crusade: he would find a way to use technology to augment human intellect.
Engelbart earned a PhD in the emerging field of computer science and joined a think tank called Stanford Research Institute (SRI), where he set up a group named the Augmented Human Intellect Research Center, later called simply the Augmentation Research Center or ARC. The group developed its On– Line System (referred to as NLS). Then, with his funding getting tight, Engelbart booked a last–minute slot at the conference to present NLS to the world.
It was a huge gamble, and a remarkable technical feat to coordinate cameras and audio feeds over a leased line, with multiple simultaneous operators miles apart in different locations. But the demo gods were gracious and, amazingly, it went without a glitch.
The demo wasn’t about the computer he was using, but in one crucial respect it was about the hardware: Engelbart was using a mouse. If he is remembered today, it is as the inventor of the mouse — he was granted US Patent number 3,541,541 for his “X–Y position indicator for a display system.” This makes it clear how the device worked: it had two wheels inside, mounted at right angles to each other, each attached to a potentiometer to measure how far it turned. When the user moved the mouse, one wheel would detect movement in the X direction and the other movement in the Y, making it easy to calculate where to move the on–screen marker relative to its starting point.
Instead of the wheels themselves making contact with the table, later evolutions of the mouse would
have a roller ball (an innovation developed by one of the 17 members of the ARC team, Bill English, who constructed the first mice with Engelbart), but the basic principle was the same, at least until the optical mouse was invented.
In the demo, Engelbart was apologetic about the name of his pointing device: “I don’t know why we call it a mouse. It started that way and we never did change it.” You only have to look at his mouse, though, to see why — in its early development the cable attached at the bottom, at the opposite end to the buttons, so it looked just like a small creature (admittedly rather angular in shape) with two eyes and a nose at one end and a tail at the other.
The cable was moved to the button end when testing established that it tended to get caught up in the user’s sleeve and generally get in the way, but at one point in the demo, when Engelbart handed over to another of his team, Don Anders, working at SRI in Menlo Park, you can clearly see that Anders was using the old–style mouse with its cable at the bottom.
Other terms used by the ARC team have not stood the test of time. For example, they call the on–screen position marker a “bug” (and joke about “bug fights” when there are two users sharing a screen), but of course a computer bug is a completely different thing today. The NLS position marker, more formally referred to as a “tracking spot”, was a relatively crude indicator too, not an elegant arrow like the familiar modern pointer, and it did not transform into a text insertion cursor or anything else.
A window to the future
Of course, a pointing device is… er, pointless without a graphical environment, and that’s the other side to Engelbart’s achievement. He not only invented the mouse, but the windowed interface that it operated upon too. During the demo, his face was projected on one side of the huge screen at the front of the auditorium and his CRT’s display on the other, but that was just for show. NLS’s interface itself could be separated into distinct zones — at one point, it had a video feed of a remote co–worker in the top–left quadrant, the area for typing commands next to it, and the shared document they were both viewing in the half–screen below.
Engelbart began the demo by typing words on a white screen, deleting some, then copying and pasting some. He was interactively manipulating
‘documents’ on screen. Then he went further. Having pasted a block of text several times to create more text than would fit on his screen (which, significantly, it did not seem possible to scroll down), he hid the text of each paragraph except for its first line (much as modern word processors might switch from full text to outline view). He wrote a shopping list and reordered it to group similar items (types of fruit together, canned goods, and so on), and then toggled between showing only category headings and the full list. He recategorized items, then showed and hid numbering, which appeared to be added automatically. He displayed a diagram of the shops he needed to visit, and selected one to view the items on his shopping list relevant to that location. Fast forward to 2019, and of course, we now call this hyperlinking.
He returned to this kind of hyperlinking later in the demo, showing how easy it was to navigate a nested hierarchy of key terms, with selecting each term opening up a linked branch or a new level of further information. The ARC team used the system for its own “planning, designing, debugging and documenting,” but Engelbart emphasized how structuring information like this was a way of organizing thought and “representing information structures that linear text cannot represent.”
Some years later, computing pioneer Alan Kay noted that there were precedents for some of this, including multiple windows and display text editing on a bitmapped screen, but the NLS demo pulled it all together and showed it all in a way that fully worked. It wasn’t Vannevar Bush’s desk–sized memex with levers and rows of buttons, and it didn’t use microfilm storage, but its windowed interface accomplished the aim of the memex’s multiple displays, and it delivered the new relationship with information that Bush had previously only dreamed of.
And that co–worker in the corner of the screen? NLS supported remote collaborative working — Engelbart allowed a colleague at another terminal to view his document at the same time, and move a second “tracking spot” on the screen; they set up audio coupling so that they could talk to each other, as well as point to what they were discussing. It wasn’t until decades later that this kind of simultaneous collaborative working was available, as in shared online iWork apps or Google Docs, although in NLS only one user at a time had control of the “chalk on this blackboard.”
The road to the Mac
So how did “the mother of all demos” lead to the Macintosh? The route was circuitous. ARC’s funding
was precarious, and NLS had got as far as it did only because Engelbart had gone out and secured money for it. Initial development of the mouse had been funded by NASA’s Bob Taylor, and more came from the Advanced Research Project Agency (ARPA) at the US Department of Defense. JCR Licklider, then the director of the Information Processing Techniques Office within ARPA, had himself published a paper some years earlier championing the prospect of human–computer interactivity. In 1965, Bob Taylor succeeded Licklider at ARPA and channelled more funds to Engelbart. As Steven Levy puts it, “For eight years the money flowed — modestly by defense
standards, topping out at about $25 million a year — until some persnickety senators forced the agency to limit its spending only to projects with specific military applications.”
By this time, Taylor had established a new think tank for the Xerox Corporation, the Palo Alto Research Center or Xerox PARC, down the road from SRI. He recruited a number of the most innovative computer scientists in the country, including some of ARC’s best and brightest, and PARC took up the baton of developing mouse–driven computing. Then, in December 1979, PARC received two visits by key personnel from an up–and–coming computer company named Apple, and the rest, as they so often say, is history.
Or rather, it’s mythology. There’s a persistent myth that Apple stole the user interface from PARC — Steven Levy himself portrays it as an audacious “daylight raid,” but Levy’s own account continues to then document the many technical advances and innovations that Apple added to the mix. For example, Xerox’s unreliable three–button mouse, derived from Engelbart’s, which cost $300 to make but was always breaking down, was simplified by Apple to a one– button device that cost just $15 but proved robust and reliable. It made the mouse both easier to learn to use, as user testing proved, and commercially viable.
The truth is, both the Macintosh project and the Lisa (the Mac’s high–end predecessor) were well under way before the PARC visit; the Mac was always conceived as having a ‘user–friendly’ interface with a bitmapped display, graphics capabilities, and some kind of
graphical input device; and a number of Apple engineers were already familiar with PARC, its work and technologies such as the mouse — some had worked at PARC, and others had even been at SRI with Engelbart. (You can see the detailed rebuttal of the myth at bit.ly/parc-visit.)
From ARC to Mac
It’s fair to say, though, that Apple did draw inspiration from what it saw at PARC, and it is clear that there’s a distinct line of descent running from Engelbart’s ARC to Xerox PARC and then to the Lisa and the Macintosh. As Levy puts it: “The Engelbart–PARC paradigm was the bedrock from which Apple was to construct the future, first with the Lisa, then the Macintosh.” NLS was a proof of concept, a first working implementation of an entirely new way of computing.
Of course, not everything in NLS took off. First, as well as the mouse and keyboard, Engelbart used a five–keyed keyset; pressing keys or combinations of keys on it entered various commands. The problem was it took too much time to master. Second, in the early part of the demo, NLS seems to respond to commands with various beeps and sounds — a regime that would have made offices very noisy places. And third, in NLS the mouse was just a pointing device with which you indicated an on–screen item you wanted to operate on; commands then had to be typed in, and in a special area of the screen at that. It was PARC that added the concept of menus, and windows that could overlap rather than being fixed screen zones. And again, Apple refined things further still, adding such key aspects as pull–down menus, direct manipulation of windows and icons, drag–and–drop file manipulation, and much more.
For all that, as Steven Levy expresses, “It was Engelbart who devised the first tools to propel one through cyberspace.” In his landmark 1984 demo, “With his keyboard, his keypad, and his mouse, Engelbart embarked on a journey through information itself. As windows opened and shut, and their contents reshuffled, his audience stared into the maw of cyberspace. Not only was the future explained, it was there, as Engelbart piloted through cyberspace at hyperspeed.” We’ve come a long way in the last 50 years, but it was Doug Engelbart who first showed, and therefore paved, the way.